131 / 162
129
6
DISCUSSION
Clinical research has shown a correlation between the DR and speech perception that endorses the need for
setting the proper fitting levels for cochlear implants [Blamey et al., 1992;Pfingst et al., 2004;Pfingst and
Xu, 2005;van der Beek et al., 2015]. The aim of the present study was to determine whether radiological
data provided additional information for setting the speech-processor map levels. Therefore, the intra-scalar
positions of individual cochlear implant electrode contacts was determined according to the CT scans of
130 post-lingually deafened subjects, and their correlations with the clinical fitting levels along the array
were studied. Interestingly, the fitting levels did not show any correlation with the distance between the
electrode contacts and the modiolus. Speech perception was not significantly correlated with the insertion
depth or the distance from the electrode array to the modiolus. Furthermore, the increase in levels at the
basal end of the array was not significantly correlated with the subjects’ mean stimulation level, duration of
deafness, age at implantation or time since implantation.
Clearly, the angular location of an electrode contact affects its T-level (Figure 3), and whether the additional
knowledge obtained from a radiological analysis would help with fitting the patient was considered. A
previous study [van der Beek et al., 2015] showed that T-levels could be fitted by determining the T-level at
one electrode contact and applying a closed-set formula for the T-level profile based on group data. Adding
radiological data to such a model might further increase the applicability of this approach for clinical
programming. Two sub-groups were created that included the 25% of subjects with the most shallowly
inserted electrodes and the 25% with the most deeply inserted electrodes (Table 6). Figure 6 shows that
the insertion depth of the array significantly affected the T-level profile. This effect was further analyzed by
adding the insertion depth to the population-based predictive formula for the levels [van der Beek et al.,
2015]. Although this process yielded a significant parameter, it did not increase the predictability of the
levels (data not shown).
The lack of correlation between the T-level and the distance to the modiolus (Figure 4) precluded the
use of this radiological parameter when setting the fitting levels. However, it is important to note that
the electrode-to-modiolar distance measurements in this study were all determined for an electrode that
was designed to be positioned at the lateral wall. Furthermore, the studies in which perimodiolar arrays
were compared with straight (lateral) arrays did not demonstrate an unequivocal correlation between the
modiolar-electrode distance and the levels. Some studies showed that perimodiolarly positioned electrodes
were associated with lower levels [Saunders et al., 2002], whereas others did not find such an effect for the
distance to the modiolus [Huang et al., 2006;Marrinan et al., 2004;van der Beek et al., 2005;Long et al.,
2014]. Gordin et al. showed that lateral packing of the cochleostomy decreased the basal ECAP thresholds
and increased the mid-array thresholds [Gordin et al., 2010]. The packing most likely decreased the basal
distance to the modiolus and increased this distance for the mid-array. This hypothesis is consistent with
the distances to the modiolus that were observed in our study population (Figure 2), in which an extended
round-window approach with lateral packing was applied. Nevertheless, despite the smaller distance from